Key Takeaways
- • GMO's systematic global macro strategy is currently short Korean and Taiwanese equities, positioning against markets that have surged 127% and 66% respectively in US dollar terms over the past twelve months, driven primarily by semiconductor demand from AI infrastructure buildout.
- • The firm's portfolio maintains a contrarian AI-themed positioning: long US equities, short US technology, short Korea, and short Taiwan—reflecting a thesis that valuations in AI-beneficiary markets are unsustainably elevated relative to long-term fundamentals.
- • Martin Emery, portfolio strategist for GMO's systematic global macro team, cites historical research linking elevated corporate investment confidence and optimistic CEO commentary to subsequent lower earnings surprises and reduced equity returns, suggesting current AI spending may replicate patterns of past investment booms preceding recessions.
- • Korea and Taiwan's outsized equity performance has been concentrated among index heavyweights Samsung Electronics, SK Hynix, and Taiwan Semiconductor Manufacturing Company—all direct beneficiaries of the semiconductor supply chain's expansion during the AI buildout.
- • GMO assesses Korean and Taiwanese equities as "particularly expensive," with long-term value forecasts trading below the cash rate, forming the fundamental basis for the short positioning across both markets.
Summary
GMO's systematic global macro strategy has established short positions in Korean and Taiwanese equity markets as part of a broader thesis that artificial intelligence investment spending mirrors historical patterns of corporate overinvestment that precede market corrections and reduced returns. According to Martin Emery, portfolio strategist for GMO's systematic global macro team, the firm does not take direct stock positions but is substantially influenced by the ongoing AI buildout in its current macro allocation decisions.
The strategy's positioning reflects a contrarian approach to the recent AI-driven rally. Over the past twelve months, Korean equities have appreciated 127% in US dollar terms and Taiwanese equities 66%, making both among the region's best performers. This performance has been heavily concentrated in semiconductor-related equities, particularly Samsung Electronics, SK Hynix, and Taiwan Semiconductor Manufacturing Company, which have served as primary suppliers and direct beneficiaries of global AI infrastructure investment. GMO's response is to maintain long positions in US equities while shorting both US technology and the Asian semiconductor markets that have driven regional outperformance.
Emery's conviction rests on empirical research examining the relationship between corporate investment cycles and equity returns. He cites studies demonstrating that elevated levels of corporate investment and positive CEO commentary during quarterly earnings calls serve as reliable negative predictors for subsequent market returns and earnings surprises. Applied to the current AI spending environment, this framework suggests that management overconfidence driving accelerated capital expenditure will ultimately result in lower aggregate earnings growth and depressed equity valuations. "In other words, higher manager confidence precedes greater investment growth and lower earnings surprises, in aggregate," Emery stated.
The valuation assessment underpins the short thesis. GMO characterizes both Korean and Taiwanese equities as "particularly expensive" relative to fundamental value, with long-term price forecasts positioned below cash rates—indicating the firm's models suggest these markets offer negative real returns over extended horizons. This positioning aligns with the systematic macro strategy's discipline on valuation and timing metrics, contrasting sharply with the momentum-driven appreciation that has characterized both markets during the AI investment cycle.
The strategy's positioning reflects a contrarian approach to the recent AI-driven rally. Over the past twelve months, Korean equities have appreciated 127% in US dollar terms and Taiwanese equities 66%, making both among the region's best performers. This performance has been heavily concentrated in semiconductor-related equities, particularly Samsung Electronics, SK Hynix, and Taiwan Semiconductor Manufacturing Company, which have served as primary suppliers and direct beneficiaries of global AI infrastructure investment. GMO's response is to maintain long positions in US equities while shorting both US technology and the Asian semiconductor markets that have driven regional outperformance.
Emery's conviction rests on empirical research examining the relationship between corporate investment cycles and equity returns. He cites studies demonstrating that elevated levels of corporate investment and positive CEO commentary during quarterly earnings calls serve as reliable negative predictors for subsequent market returns and earnings surprises. Applied to the current AI spending environment, this framework suggests that management overconfidence driving accelerated capital expenditure will ultimately result in lower aggregate earnings growth and depressed equity valuations. "In other words, higher manager confidence precedes greater investment growth and lower earnings surprises, in aggregate," Emery stated.
The valuation assessment underpins the short thesis. GMO characterizes both Korean and Taiwanese equities as "particularly expensive" relative to fundamental value, with long-term price forecasts positioned below cash rates—indicating the firm's models suggest these markets offer negative real returns over extended horizons. This positioning aligns with the systematic macro strategy's discipline on valuation and timing metrics, contrasting sharply with the momentum-driven appreciation that has characterized both markets during the AI investment cycle.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.